基于TM/ETM+数据的沈阳市各区城市热岛特征

生态学杂志

基于TM/ETM⁺数据的沈阳市各区城市热岛特征

王宏博¹,李丽光^1**,赵梓淇¹,蔡福¹,武晋雯¹,许申来²,姜鹏^3

（¹中国气象局沈阳大气环境研究所，沈阳 110166； ²北京清华同衡规划设计研究院，北京 100085； ^3辽宁省气象学校，沈阳 110016）

出版日期:2015-01-10 发布日期:2015-01-10

Urban heat island variation of each district in Shenyang based on TM/ETM⁺ data.

WANG Hong-bo¹, LI Li-guang^1**, ZHAO Zi-qi¹, CAI Fu¹, WU Jin-wen¹, XU Shen-lai², JIANG Peng³

(¹Institute of Atmospheric Environment, China Meteorological Administration, Shenyang 110166, China; ²Beijing Qinghua Tongheng Planning and Design Research Institute, Beijing 100085, China; ³Liaoning Meteorological School, Shenyang 110166, China)

Online:2015-01-10 Published:2015-01-10

摘要/Abstract

摘要：

应用2001年8月和2010年8月Landsat TM/ETM⁺数据，计算沈阳市区及三环内各区的地表热岛强度（SUHI），根据土地利用数据和热岛源汇特征提取源汇信息，分析SUHI与土地利用类型、热岛源汇面积及边界长度、归一化植被指数（NDVI）、归一化建筑指数（NDBI）、改进的归一化差异水体指数（MNDWI）的关系。结果表明：2001年沈阳市区的中等热岛强度以上区域主要集中在三环内和苏家屯区，强热岛地区主要出现在铁西区和皇姑区；2010年中等以上热岛强度地区，与城区发展相一致，主要向西南，南面扩展，弱热岛、中等强度和强热岛地区有较大幅度的增加；城市热岛强度与NDVI存在显著的负相关，与NDBI存在显著的正相关；热岛源在区域内的面积比重与中等强度以上热岛存在较显著的相关关系，源汇边界长度与弱热岛和中等热岛存在较显著的相关关系。

关键词: BP神经网络模型, TDR-300, 微波遥感, 全极化SAR, 森林地表土壤含水率, 多元线性回归模型

Abstract:

Base on the Landsat TM/ETM⁺ data of August in 2001 and 2010, the surface urban heat island (SUHI) intensity in each district within the thirdcircle freeway of Shenyang was calculated. The data of land use types were used to obtain the information of SUHI source and sink, and the relationship between SUHI and land use types, sourcesink area and boundary length, normal differential vegetation index (NDVI), normal differential built-up index (NDBI), modified normal differential water index (MNDWI) were analyzed. Results show that, in 2001, the areas of medium or high SUHI intensity were mainly within the third circle freeway of Shenyang and Sujiatun district, and the areas of strong SUHI intensity were in Tiexi and Huanggu districts. In 2010, medium or high SUHI areas mainly extended towards the south and southwest, and light, medium and high SUHI areas increased significantly as compared with that in 2001. There was a significant negative correlation between SUHI and NDVI, and a significant correlation between SUHI and NDBI. The proportion of heat island source area to total urban area was significantly correlated with the medium or high SUHI, and the sourcesink boundary length was significantly correlated with the light and medium SUHI.

Key words: BP neural network model, multivariate linear regression model, TDR-300, moisture content of forest surface soil, Quad-pol SAR, microwave remote sensing

王宏博1,李丽光1**,赵梓淇1,蔡福1,武晋雯1,许申来2,姜鹏3. 基于TM/ETM⁺数据的沈阳市各区城市热岛特征[J]. 生态学杂志.

WANG Hong-bo1, LI Li-guang1**, ZHAO Zi-qi1, CAI Fu1, WU Jin-wen1, XU Shen-lai2, JIANG Peng3. Urban heat island variation of each district in Shenyang based on TM/ETM⁺ data.[J]. cje.